Nature and Measurements of Torque Ripple of
Permanent-MagnetAdjustable-SpeedMotors
John S . Hsu (Htsui), Brian P. Scoggins, Matthew B. Scudiere, Laura D. Marlino, Donald J. Adams, Pragasen Pillay
Senior Member
Senior Member
Oak Ridge National Laboratory
P.O. BOX2003, K-1220, MS 7280
Faculty Research
Participant
University of New Orleans
Oak Ridge, TN 37831-7280
Abstract: Torque ripple of permanent-magnet motors can be
classified into four types depending on the nature of their
origin. The four types are pulsating torque, fluctuating
torque, reluctance cogging torque, and inertia and mechanical
system torque.
buried magnets [ 1,2, 31,
The cogging torque of a PM motor is produced by nonuniformly distributed air-gap permeance associated with the
teeth and the permanent magnets. Skewing either the teeth
or the magnets is one approach for cogging-torque
reductions [3]. It is also possible to reduce the cogging
torque by properly choosing the ratio of magnet width to
slot pitch without skewing [4]. Other approaches, such as
tapering the magnet edges, may also reduce cogging
torques.
Pulsating torques are inherently produced by the
trapezoidal back-emf's and trapezoidal currents used in
certain permanent-magnet adjustable-speed motors. The
torque ripples caused by pulsating torques may be reduced by
purposely produced fluctuating counter torques.
Air-gap torque measurements are conducted on a sample
motor. Experimental results agree with theoretical
expectations.
The air-gap torque in Newton meters of a PM motor is
I. INTRODUCTION
i e +ie +ie
T= aa b b c c
(speed 1
Torque ripple of adjustable speed permanent-magnet
(PM) motors have adverse effects in many applications.
For example, low surface finish roughness for machine tool
drives depends on motor torque quality. For propulsion and
vehicle motors, quietness and smoothness are strongly
required.
23
T
-
60
where ea, eb, and ec are the back electromotive forces
(emf's) induced i n the armature windings by permanentmagnets, while ia, ib, and ic are the armature currents. The
rotor speed is given in revolutions per minute (rpm).
The severity of torque ripple increases at low speeds. In
addition, the effect of the moment of inertia that tends to
absorb shaft speed variations is less significant at lower rate
of speed fluctuations. A smooth air-gap torque is
particularly desired at low speed.
Various data-signal-processing (DSP) approaches are
possible for controlling the constant air-gap torque
presented by equation (1). When it is assumed that ea, eb,
and ec are not changed under various loads, either the dclink current or the armature currents can be adjusted to
change ia, ib, and ic. The latter approach [5] may reduce
torque ripples caused by commutation.
There are various inverter and motor schemes for
permanent-magnet adjustable-speed drives [ 1-6, 11-17].
Most existing literature studies motors having surfacemounted magnets. A minor portion of the literature studies
When saturation effects are considered in PM motors,
Chen, Levi, and Pelka [6] suggest that the field excitation
voltages of PM alternating-current (ac) motors can be
considered constant at their open circuit values for all points
of operation. On the other hand, Ostovic and Osijck [ 11 use
a flux tubes method to evaluate transient states in a PM
machine. The nonlinear characteristics of the induced
voltages and the torque are calculated.
Most existing literature on the torque ripple of PM
motors appears to be theoretical investigations. Few studies
conduct tests, and laboratory torque measurements are
generally not included.
MSTRlBUTlON OF THIS DOCUMENT IS UNLlMff
1
1
S
-
~
~
~
~
DISCLAIMER
?'his nport was prepared as an account of work sponsored by an
agency of the
United States Government Neither the Unit& States Government nor any agency
thereof. nor any of their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility for the accuracy, completeness, or Wfulness of any infomation, apparatus, product, or proass d i s c i d , or rrprescnts
that its use would not infringe privately owned rights. Reference herein to any spe-
cific commercial product. ptoctss. or lervicc by trade name, trademark, manufacturer. or otherwise does not necessarily constitute or imply its endorsement, m m mendation, or favoring by the U n i d States Government or any agency thereof.
T h e views and opinions of authors exprrsscd herein do not n d l y state or
reflect those of the United States Government or any agency thereof.
DISCLAIMER
Portions of this document may be illegible
in electronic image products. Images are
produced from the best available original
document.
harmonic component shown in Table 1 represcnts a 20-pole
winding connected in series with the 4-pole fundamental
winding. For three phase windings, the locations of the 20pole harmonic windings are in negative sequence. With
reference to their winding factors, the significance of the
20-pole harmonic windings viewed from the winding
terminals is 7% of the fundamental winding component.
This paper looks into the origin of four types of torques
that are components of torque ripple of adjustable-speed
PM motors. Experimental results obtained from laboratory
air-gap torque measurements agree with theoretical
expectations.
11. SPATIAL
AND TIME
HARMONICS
The armature windings and the permanent magnets of
PM motors generally contain spatial harmonics, and their
armature currents have time harmonics.
B. Spatial Harmonics of Permanent-Magnets Flux
The rotor of the experimental motor has four surfacemounted permanent magnets The pole span of each
magnet is 140 electrical degrees. The spatial harmonics of
a trapezoidal wave having a 140 electrical-degree-wide
plateau are:
A. Winding Spatial Harmonics
The motor windings of a phase are normally situated in
particular slots. For three-phase windings, each phase
winding is located 120 electrical degrees apart from the
others, The fundamental a-, b-, and c-phase-winding
locations follow a sequence that is in the same direction as
the normal shaft rotation. This sequence is called the
positive sequence.
Orderof harmonics:
11
0.07
0.07
0.50
0.93
9
7
11
C. Back emf Induced by Permanent Magnets
When the spatial harmonics of the PM rotate in the air
gap at the rotor speed, the emf magnitudes and frequencies
induced in the stator winding depend on the individual
winding spatial harmonics. An emf of a particular
harmonic order is induced only when both the stator
winding harmonic component and the rotor PM flux spatial
harmonic component are of the same order. Table 2 shows
the relative frequencies of the back emf's of the sample
windings given in Table 1 with the above sample spatial
harmonics of PM flux.
Back emf's induced by permanent magnets
Order of
Relative
Relative
harmonics
amplitude
fresuency
1
100%
Io
3
17%
30
5
1%
50
7
<l%
70
Double-layer winding, 60-degree phase belt
Coil pitch=l - 6
0.50
5
The spatial harmonics of the rotor PM flux are rotating
in the positive direction at the rotor mechanical speed.
No. of p o l e s 4
No. of stator slots=24
3
5
7
9
3
Relativeamplitudes: 100% 31% 13% 4% 0% 2%.
If a current is flowing through the winding, the
magnetomotive force (MMF) generated is not a pure sine
wave. Each slot has a ceftain concentration of ampere-turns
that produces a discrete change of MMF from its adjacent
teeth. The M M F wave of the winding contains
fundamental and harmonic components. These components
represent the existence of winding spatial harmonics. In
practice, the winding spatial harmonics are calculated
through winding factors [7]. The values for an
experimental motor used later are listed in Table 1.
Spatial stator-winding harmonics:
I Winding factor, Kn
Orderof
harmonics, n
1
0.93
1
I
I
I Relativesequence
of three phases
Positive
Zero
Negative
Positive
Zero
Negative
9
0%
2%
11
90
1 lo
'
o-Fund. electrical angular frequency
Table 2 Back emfs induced by PM of experimental motor
D. Time Harmonics of Armature Currents
Table 1 Spatial stator-winding harmonics
Although phase windings are located i n positive
sequence, their individual harmonic windings may be
located in either positive, negative, or zero sequence, where
the negative sequence is in the opposite direction to the
positive sequence. Zero sequence means that the threephase harmonic winding components are located at the
same positions.
Currents
0
The harmonic windings having numbers of poles
proportional to their harmonic orders are connected in
series. For example, the fifth spatial stator-winding
.-
\
Fig. 1 Sample trapezoidal armature currents
2
.
..a.
L
Y
I Trapezoidal armature-currentharmonics
Order of
harmonics
1
11
Relative
amplitude
100%
0%
1%
Relative sequence
of three phases
Positive
magnitudes of phase currents in the same ratio. The sample
DSP approach discussed earlier is an example of using
torque fluctuations for torque ripple elimination.
I
In a given motor the back emf is fixed and torque and
torque ripple can only be regulated by maintaining the
proper instantaneous current in each of the three phases.
Since the operating voltage and the switching time intervals
are bounded, current regulation cannot be exact or in some
cases may not even be close to the desired value. This
occurs at the higher end of the operating speed where the
back emf is large compared to the driving voltage. When
the inverter switch is set to increase the current, the
charging voltage across the motor inductance is the driving
voltage minus the back emf. In this region the resultant
voltage can be too small to build current fast enough.
When the inverter switch is set to reduce the current, the
back emf adds to the link voltage and can discharge the
current in a phase significantly faster than the charging rate.
With a Y configuration requiring the phase currents to sum
to zero, the interaction of the three phases along with the
difference in charging and discharging makes it impossible
to completely regulate each phase current precisely at all
times i n the higher operating speed regions without
increasing the source voltage. But at the lower speeds with
the switching times bounded this higher source voltage can
cause large current fluctuations and therefore produces
significant torque ripple at the switching rate.
Ne ative
Positive
Zero
Ne ative
Table 3 Time harmonics of trapezoidal currents of Fig. 1
As an example, the time harmonics of a trapezoidal
armature current shown in Fig. 1 are listed in Table 3. The
relative sequences indicate the phasal relationships in time
among three phases.
111. NATURE
OF TORQUE-RIPPLE
ORIGINATIONS
Torque ripple has four origins, which are classified as
follows:
A. Pulsating Torques
This type of torque is usually the most common in PM
motors. Pulsating torque originates from the interaction
between stator and rotor fields. In order to produce a
pulsating torque the following conditions must be met:
1) The stator field and the rotor field have an identical
number of poles. Fields having different numbers of poles
do not interact with each other.
C. Reluctance Cogging Torque
Armature teeth present non-uniformly distributed airgap reluctance. The reluctance torque produced by teeth
and permanent magnets varies as their relative positions
change. This torque can be felt while turning the motor
shaft by hand without energizing the motor. The cogging
torque versus shaft position can be measured in a similiar
manner. Different approaches for cogging torque
reductions can be found in the literature.
2) The stator field and rotor field are rotating at different
speeds. One field may be stationary, or both may rotate in
either the same or opposite directions.
3 ) The frequency of the pulsating torque is the
difference of the rotating speeds per second of the stator
and rotor fields
For example, the three phase currents of the 5th time
harmonic given in Table 3 are in negative sequence. When
these currents go through the fundamental winding
component shown in Table 1, a 4-pole rotating field at five
times the rotor speed but i n a reversed direction is
generated. The interaction of this field with the
fundamental forward-rotating field of the rotor produces a
sixth order of pulsating torque. The torque magnitude is
proportional to the product of the rotor fundamental field
magnitude, the stator 5th order winding factor, and the
relative amplitude of the 5th harmonic of the armature
current. Other orders of pulsating torques can be evaluated
in like manner.
~
D. Inertia and Mechanical-System Torques
When the speed of the rotor is changed, a torque
proportional to the product of its moment of inertia and the
rate of speed change is generated. This torque resists the
change of speed. The moment of rotor inertia acts like a
filtering capacitor in an harmonic-prone electronic circuit.
At low speed, the rate of speed ripple is lower.
Subsequently, this torque is weaker. The mechanical
components .of the motor can be considered as a spring
system. Certain torques are generated according to thedynamic motions of the spring system.
Iv.AIR-GAPTORQUE OF THREE-LEAD,
If the steady air-gap flux distribution is spatially
sinusoidal and the armature steady input currents are
sinusoidal in time, the motor does not yield pulsating torque
components.
THREE-PHASE MOTORS
Air-gap torque equations using measurable data have
been known for several decades {8-10, 18, 191. One
familiar expression [8] is presented i n (2).
B. Fluctuating Torques
Equation (3) is rewritten [9] from (2) for the line data as
depicted in Fig. 1, where the symbols are denoted by their
Torque fluctuations can be produced by altering the
3
E
corresponding line identifications in the suffixes.
inverter
3-phase
power
source
Eddy
current
brake
where suffixes a, b, and c stand for phase values,
p is number of poles, i represents current,
pi
and y is for flux linkage
Torque [Nm]=
3
-
(iA
-
6%)
+(iC-iA)
.J [VcA - R (ic - i A )
]
df
.I[vAB- R(iA -iB)l
1
(3)
Equation (3) is valid for either Y- or delta-connected
motors, where
iA, ig, and ic = line currents
R = half of the line-to-line resistance
value.
From the definition of R, the following two expressions
are for Y and delta-connected motors, respectively.
R = phase resistance, r, for Y-connected motor
R = r/3 for delta-connected motor.
Equation (3) contains certain errors due to the following
assumptions and simplifications. First, the three-phase
leakage reactances are linear and identical. Second, the
negative sequence winding (not supply) components are
negligible. Third, the torque components produced by
sources that are not dependent on the armature winding
currents, such as permeance cogging torque in a permanentmagnet motor, are not considered. Fourth, the instantaneous
magnetic unbalances [lo] for three phases are ignored.
Fifth, the effect of dc components in the flux linkages are
neglected. Finally, the equations are not affected by minor
speed ripples.
Fig 2. Measurement setup for air-gap torque
current and voltage transducers are used. Isolated
transducers are advantageous since the signals can be
connected to a single common point. This is a great
convenience to the data acquisition system provided that
additional instrumentation errors are negligible.
A four-channel, dc to 175 MHz, digital storage scope is
used. It has a GPIB (IEEE488) input/output interface
capability. Maximum sample rate is 100 megasample per
second, per channel at 5 ms/div. Its storage capacity is
4 x 10k.
When using either three leads for Y-connected motors
without a neutral connection or three leads for deltaconnected motors, (3) can be further simplified by using
i g = - (iA + ic). The above equation can be rewritten to
the known format that uses only two line voltages, two line
currents, and one-half of the line-to-line resistance as the
input data for the calculation of the air-gap torque.
Torque [Nnl] =-
P .J7
6
Sample time differences among different channels are
neglected in this study. This can be included easily if the
specific time differences are given; however, when the time
increment for sampling is small, the error relating to the
time differences is negligible.
B. Air-Gap-Torque Measurement of an Adjusted-Speed
Brushless DC Motor
- (2'iA+iC).I[VCA - R ( i c - i ~ ) ] d f
+ (ic-id .I[-
VBA
- R (2. iA + ic)] dr
The air-gap torque includes pulsating and fluctuating
torque components that are affected by armature currents.
A 1-hp, 4-pole, 230-V, three-phase brushless dc permanentmagnet motor is fed by an adjustable-frequency inverter.
The motor is coupled to an eddy-current-disk load with
torque gauge.
(4)
v. AIR-GAP-TORQUE
MEASUREMENTS
A. Experimental Setup
The output of the inverter is extremely unbalanced due
to defective hardware inside the inverter. Since one
The experimental setup is shown in Fig. 2. Isolated
4
~
~
i:E{
HIGH LOAD
LOW LOAD
in
-20
1
982
z
1
Line c u r r e n t s
Row a.
v
v3~~~~
E8
-300
1
982
Row b.
- 20
982
v
- 300
1
982
Line c u r r e n t s
30
v
v 88
982
1
L i ne voltages
-30
1
w2
1
Li ne voltages
Nm
Row c.
1
-101
1
982
Torque waveform
1
Torque waveform
I
982
Fig. 3. Current, voltage, flux linkage, and air-gap torque waveforms (with 982 data points per cycle)
look relatively square. The waveforms under the low load
situation contain two bumps in each cycle of current. This
indicates greater higher-order harmonics. The three phase
currents are extremely unbalanced. Phase b current, which
is the negative summation of iA and ic, is nearly zero.
objective is to examine the air-gap torque measurement
method that includes the situation of an unbalanced supply,
the defective inverter is used.
Tests with high and low loads, respectively, are
conducted without decoupling the motor. The high load
refers to the highest possible load under unbalanced supply,
which is 172 watts, and the low load is 5 watts. Both tests
are at 1550 rpm.
The harmonic current components for both the high
and low loads, respectively, are listed in Table 4. The bold
face numbers indicate the harmonic values that are
obviously greater than the rest. According to the
description given earlier in section I11 B, the predicted
major frequencies of pulsating torques produced by all the
current components shown in Table 4 and the fundamental
winding component alone are given in bold face numbers in
the right column of the same table. The expected major
torque harmonic frequencies are lo, 20, and 4w. The
effects of other winding harmonic components can be
evaluated in like manner.
Measured waveforms of one cycle for the high and the
low loads are placed side by side in Fig. 3. The left two
columns are for the high load and the right two columns are
for the low load.
The first row (row a) of Fig. 3 shows the line currents
and ic. They contain dc components caused by the
defective inverter. The current waveforms under high load
iA
5
with the low load situation.
S tator-ci -ent harmonics of experimental motor
Relative I Sample
Orderof Amplitude [amps]
sequence torque
current
of three
freq. by
harmonphases
fund.
ics
winding
I
The last row (row c) of Fig. 3 illustrates the tested airgap torque waveforms calculated from (5). The harmonic
contents of the tested torque waveforms are listed in Table
5. The significant harmonic frequencies are agreeable with
those predicted frequencies given in Table 4. It is also
expected that, as shown in Table 5, the higher order of
torque harmonics under low load are more significant than
those under high load.
1
dc
I v . CONCLUSIONS
2
-
1) Torque ripple of a permanent-magnet motor can be
classified into four types of torques according to the nature
of their origin.
2) The four types are pulsating torques, fluctuating torques,
reluctance cogging torques, and inertia and mechanicalsystem torques.
3
3) Pulsating torque is generated by the interactions between
stator and rotor fields having the same number of poles, but
rotating at different speeds.
4) Pulsating torques are inherently produced by the
trapezoidal back-emf‘s and trapezoidal currents used in
certain permanen t-magnet adjustable-speed motors.
5 ) The torque ripples caused by pulsating torques may be
reduced by purposely produced fluctuating counter torques.
6 ) Reluctance cogging torque is produced even when the
motor is not energized, and can be measured when the
motor is not energized.
7) Inertia and mechanical-system torques are generated by
the dynamic motions of the mechanical components of
motor. They are affected by the driven device.
8) Air-gap torque measurements for pulsating and
fluctuating torques are conducted on a sample motor.
9) Experimental results agree with theoretical expectations.
0
d
;able4 2
and expected significant frequekies of torque
The second row (row b) of Fig. 3 presents line
voltages VBAand VCA. The voltages at the low load appear
to have more notches than the voltages under highest load.
Consequently, higher harmonic magnitudes are associated
V. ACKNOWLEDGMENTS
The authors would like to thank the Oak Ridge National
Laboratory (ORNL) for the support staff and facilities
provided for the research work. The funding provided for
the research is deeply appreciated. The authors would also
like to acknowledge the earlier study done by Mr. Anthony
Lopez [20] on brushless dc motors.
Air-gap torque harmonics of experimental motor
Torque
Amplitude [Nm]
harmonic
VI. REFERENCES
[I]
[2]
[3]
experimental motor
6
Vlado Ostovic, Istarska B. Osijck, “Computation of
Saturated Permanent Magnet AC Motor Performance by
Means of Magnetic Circuits,” IEEE Paper No. CH22723/86/0000-0794, 1986 IEEE. P.P.794-799.
Alexander Levran, Enrico Levi, “Design of Polyphase
Motors with PM Excitation,” IEEE Transactions on
Magnetics, Vol, Mag-20, No. 3, May 1984, p.p. 507515.
Jaime De La Ree, Nady Boules, “Torque Production in
Permanent-Magnet Synchronous Motors,” IEEE Paper No.
CH2499-2/87/000-0015, 1987 IEEE, P.P. 15-20.
Touzhu Li, Gordon Slemon, “Reduction of Cogging Torque
in Permanent Magnet Motors,” IEEE Transactions on
Magnets, Vol. 24, No. 6, November 1988, p.p. 2901-2903.
Renato Carlson, Michel Lajoie-Mazenc, Joao C. dos S.
Fagundes, “Analysis of Torque Ripple due to Phase
Commutation in Brushless DC Machines,” IEEE Paper No.
90/CH 2935-5/90/0000-0287, 1990 IEEE, P.P. 287-292.
Ming-De Chen, Enrico Levi, Mark Dov Pelka, “Iron
Saturation Effects in PM AC Motors,” IEEE Transactions
on Magnetics, VoI. Mag-21, No. 3, May 1985, p.p. 12621265.
A. E. Fitzgerald, Charles Kingsley, Jr., Stephen D. Umans,
mctric Machinery, p.546, McGraw-Hill Book Company,
New York, 1983.
Jing-Tak Kao, B N A L Y S S OF TRANSI E F N U
QPERATIONS OF ALTERNATING-CURBEIU
M
m, Second Edition, Science Press, Beijing,
China, Sept., 1964.
J.O. Ojo, V. Ostovic, T.A. Lipo, and J.C. White,
“Measurement and Computation of Starting Torque
Pulsations of Salient Pole Synchronous Motors,” IEEEPES
1989 Summer Meeting, Long Beach, California, July 9-14,
1989. Paper No. 89 SM 756-8 EC.
J.S. Hsu, H.H. Woodson, and W.F. Weldon, ”Possible
Errors in Measurement of Air-Gap Torque Pulsations of
Induction Motors,“ IEEE Transactions on Energy
Conversion , (Paper No. 9 1 SM 390-5 EC).
Jaime De La Ree, Jaime Latorre, “Permanent Magnets
Torque Considerations,” IEEE Paper No. 88CH25650/88/0000-0032, 1988 IEEE. P.P. 32-37.
Eljas G. Strangas, Tuhin Ray, “Combining Field and Circuit
Equations for The Analysis of Permanent Magnet AC Motor
Drives,” IEEE Paper No. 88CH2565-0/88/0000-0007, 1988
IEEE, P.P. 7-10.
H. R. Bolton, Y. D. Liu, N. M. Mallinson, “Investigation
into A Class of Brushless DC Motor with Quasisquare
Voltages and Currents,” IEE Proceedings, Vol. 133, Pt. B,
No. 2, March 1986, p.p. 103-111.
H. R. Bolton, R. A. Ashen, “Influence of Motor Design and
Feed-Current Waveform on Torque Ripple in Brushless DC
Drives,” IEE Proceedings, Vol.131, Pt. B, NO. 3, May 1984.
Yoshihiro Murai, Yoshihiro Kawase, Kazuharu Ohashi,
Kazuo Nagatake, Kyugo Okuyama, “Torque Ripple
Improvement for Brushless DC Minature Motors,” IEEE
Paper No. CH2499-2/87/000-0021, 1987 IEEE, p.p. 21-26.
Tomy Sebastian, Gordon R. Slemon, “Operating Limits of
Inverter-Driven Permanent Magnet Motor Drives,” IEEE
Transactions on Industry Applications, Vol. 1A-23, No. 2,
MarcNApril 1987, p.p. 327-333.
S. Funabiki, T. Hirnei, “Estimation of Torque Pulsation due
to The Behaviour of A Convertor and An Inverter in A
Brushless DC -Drive System,” IEE Proceedings, Vol. 132,
Pt. B, NO. 4, July 1985, P.P. 215-222.
[18] John S. Hsu (Htsui), “Monitoring of Defects in Induction
Motors Through Air-Gap Torque Observation,” Paper No.
94A11, (EMC94-3I), IEEE-IAS Annual Meeting, Denver,
Colorado, USA, October 2-7, 1994.
[I91 J.S. Hsu (Htsui), A.M.A. Amin, “Torque Calculations of
Current-Source Induction Machines Using the
1-2-0 Coordinate System,” IEEE Transactionson Industrial
Electronics, vol. 37, no. 1, February, 1990, pp. 34 - 40.
[20] Anthony Lopez, “Brushless DC Motor,” Study Report, Oak
Ridge National Laboratory, 1994.
John S. Hsu (Htsui) (SMIEEE) is a senior staff scientist
of the Digital and Power Electronics Group, Oak Ridge
National Laboratory. He worked for Emerson Electric
Company, Westinghouse Electric Corporation, and later for
the University of Texas at Austin. Dr. Hsu has published
many technical papers and holds seven patents in the area
of rotating machines and power electronics.
Brian P. Scoggins is currently completing a one-year stay at Oak Ridge National Laboratory. He is a senior
attending the University of Southern Indiana majoring in
Electrical Engineering Technology.
Mr. Scoggins is a member of the US1 Presidential
Scholar’s program and the Tau Alpha Pi honor society as
well as a 1990 National Merit Scholarship Finalist, high
school valedictorian, and former page for the Indiana State
Senate.
Dr. Matthew Scudiere holds degrees in Physics from
Kent State University and Nuclear Engineering from the
University of Virginia. He has worked as a research
scientist and inventor for most of his career in areas such as
power electronics, permanent magnet motor design, and
image and signal processing, and has 20 years’ experience
in instrumentation design and control. Matt has several
awards from Martin Marietta for inventions and special
achievements, and has several patents currently being
processed as well as approximately 15 refereed papers
published.
Laura D. Marlin0 received her Bachelor’s degree in
Electronics Engineering from the University of New
Mexico in December of 1982 and her Master’s from theUniversity of Tennessee in 1991. Her industrial experience
includes employment with Teledyne Camera Systems of Arcadia, California in 1983 and Honeywell Marine and
Defense Systems in Albuquerque, New Mexico from 19831989. She is currently a staff engineer with the Power
Electronics and Digital Systems Group at the Oak Ridge
National Laboratory in Oak Ridge, Tennessee.
Donald J. Adams leads the Digital & Power Electronics
Group at the Oak Ridge National Laboratory where he has
been employed since 1977. He received a B . S . in
Mechanical Engineering from the University of Mississippi
in 1973 and a M.S. from the University of Tennessee in
1977. His work experience primarily includes process
engineering, rotor dynamics, and power electronics. Don is
a registered professional engineer in Tennessee.
Pragasen Pillay (SMIEEE) is an Associate Professor at
the University of New Orleans, having obtained the Ph.D
from VPI&SU i n 1987 while funded by a Fulbright
Scholarship. He is active in the IEEE, and he spent the
summer of 1994 at the Oak Ridge National Laboratory as a
faculty research participant.
-
=
~
~
~
T h e submitted manuscript has been authored by O a k Ridge National
Laboratory, Oak Ridge, Tennessee 37831-7280,managed by Manin Manetta Energy
Systems, Inc. for the U. S. Department of Energy under contract No. DE-ACOS8401321400. Accordingly, the U. S. Government retains a nonexclusive, royalty-free
license to publish or reproduce the published form of this contribution, or allow
others to do so, for U. S. Government purposes.
7
-~